Forks in the Road: Rethinking Modeling Decisions that Defined Teaching and Practice of Geotechnical Engineering

Forks in the Road: Rethinking Modeling Decisions that Defined Teaching and Practice of Geotechnical Engineering

Presenter: Prof. Rodrigo Salgado
Title: Forks in the Road: Rethinking Modeling Decisions that Defined Teaching and Practice of Geotechnical Engineering
Launching Date & Time: June 23 2020 14:45pm GMT

Session Question & Answer [CLOSED]

Marina Pantazidou

Ambrose (expert on Education) says:
“What we believe to be accurate (truthful, factual) is problematic when it is not accurate (truthful, factual). In part, that’s because if we believe it to be accurate and we own it, it is really hard to undo the knowledge.”
Salgado (expert on Geotechnical Engineering) says:
“From the point of view of teaching, students who learn the wrong concepts, later on will have difficulty unlearning them.”
Both say the same thing: START with the correct concept, because undoing a wrong concept may be impossible.
Cohesion may be like Santa Claus. It is not bad for children to believe in Santa Claus, as long as they grow out of this belief when older. Easy explanations for wrong concepts are extra attractive exactly because they are easy (like conspiration theories). Why abandoning them?
But I may be wrong. As a TC306 representative, I commit to gather evidence as to whether cohesion helps or hinders our understanding of soil behavior.

Charles MacRobert

This question is directed towards Prof Salgado.

Firstly, thank you for a very interesting presentation. It has been said, “The fundamental condition to be satisfied is that the method of analysis must be matched by a compatible standard of site investigation and construction control on site. If this condition is not satisfied, then at best our designs are likely to be uneconomic, at worst, we will be in danger of producing expensive structural failures.”

How to we balance the complexity of what we teach with the realities of practice especially in developing countries. Rarely do we have enough site investigation data to use anything but the simplest of strength models. Should we not teach tools that are compatible with the level of practice students will encounter? Are you suggesting we focus more on concepts/mechanisms and less on theoretical models ? Further, how do we balance this with the fact that geotech is often taught in a few courses at undergraduate level?

Rodrigo Salgado

Thank you, Dr. MacRobert, for this interesting question.

You will notice that, in the lecture, I have focused on identifying concepts that are demonstrably incorrect or models that yield results that are demonstrably inaccurate or miss an important aspect of soil response. It follows that, even if data from site investigation are not of sufficient quality or quantity, you still don’t want to match that to models or design methods that are not grounded on solid science. In other words, one mistake does not justify another. You can see from this that I am not necessarily in agreement with the quote that you included in your question. I don’t believe that we should lower the standards on one end because they are low on the other.

Naturally, insufficiency of data or their low quality will limit what we can do on the analysis side of the problem. So your question raises another important point: we should, whenever possible, teach students what a good site investigation program should be. As an example, consider the CPT versus the SPT. Any data insufficiency, however, would not be, in my view, a reason to use equally low-quality methods of analysis and design.

As to the second part of your question – which addresses what we should be teaching students in the limited number of hours devoted to geotechnical engineering at the undergraduate level – I would guide myself by the Bible: “And with all thy getting get understanding.” Proverbs 4:7. I think that it is tedious for students when we teach them correlations, charts and methods of calculation without the all-important underlying science. You don’t need to and probably should not delve too deep into constitutive modeling at that level. But you can teach students about the real sources of shear strength – friction and dilatancy – and use some simple model (I give the Bolton (1986) very simple model as an example in the lecture) to bring it all together for students.

As I also stated in the lecture, the education of today is the practice of tomorrow, so we should never tailor our teaching to match practices that are lacking in one way or another.